There are no structure modifying treatments in osteoarthritis (OA) in part because the proof of structure modification requires large scale long term studies that demonstrate treatment effects on cartilage morphology or other joint structures. Also, changes in cartilage morphology may be irreversible. If biomarkers for treatment response were available, promising treatments could be tested efficiently and might be evaluated when disease is reversible. Without ways of detecting treatment effects, OA treatment development will likely continue to be delayed. To develop biomarkers for treatment response, biomarkers need to be tested using an effective treatment. Bariatric surgery (BSX) produces dramatic improvements in OA symptoms and likely results in stability of cartilage matrix or even improvement. With loss of 100 pounds over one year on average, in many but not all patient?s knee pain resolves. The assumption behind the planned work is that the weight loss that accompanies BSX constitutes a human model of an effective knee OA treatment which either stabilizes or leads to improvement in knee structure. That model will allow us to explore how detect this improvement in structure which might exemplify structure modification and might allow other treatments to be evaluated over a short-term basis. Structural changes in cartilage that reflect abnormal cartilage can now be imaged using several different approaches including T1rho and T2 mapping. Bone marrow lesions which show bone damage histologically and occur in regions of increased loading could improve with weight loss and could be a biomarker also. The overall goals of this project are to study massive weight loss from BSX and its effects on knee pain and structural pathology in knees. The specific aims are: 1) To determine whether the improvement in knee pain in those experiencing weight loss after BSX is less likely in those with specific structural findings. 2) To characterize MRI changes before and one year after massive weight loss and in comparably obese persons not undergoing BSX. We will examine changes in imaged structure of cartilage using T1rho, T2 maps and bone marrow lesion volume.